Section 1: Pressure Volume Curves

1.1: Variation in TLPs

Look at just one species and timepoint: ABCO and fall

In fall and spring:

  • fall both dates combined ()

See if sig. difference in TLP:

1.2: Try to predict Mpa from LFM with just one species

  1. Get datasets in order:
  • select for just the species and date we are interested in
  • remove any anomalous curves
  • pull the TLP as a value
  • make two sub-datasets: one for all points above the TLP, one for all points below the TLP

For fall PV curves, the mean TLP = -2.15, sd = 0.21

Subset data based on TLP:

2. For points below the TLP, there should be a linear relationship between -1/Mpa and LFM

2.1. Try adding in other predictors:

This is the model that dredge() likes the best:

VIFs: #### 3. For points above the TLP, there should be a linear relationship between Mpa and lfm

We have some linear models that should be able to predict Mpa from LFM. Now, to test it out…

3.1. Try adding in other predictors:

This is the model that dredge() likes the best:

VIFs:

Sections 2-3: Can we predict actual values using the models built from the PV curves?

2.1: Flam-curve values

  • Read in flam curve data

  • Use predict() to see if we can use PV curve models to get values collected during flammability curve testing.

We’re just interested in ABCO, so just look there:

2.2 Predict below TLP:

2.2.1 Simple model - below:

2.2.2. Complex model - below:

Since we don’t have Po for the flam. curve data, we’ll have to do the best we can and only use swc

####2.2.3. Simple model - above:

2.2.4 Complex model - above:

3.1: Field data

This data is from measurements of predawn and midday Mpa, and midday measured LFM. SWC is equivalent to the wettest lfm measured for each invidual divided by 100 to get grams water/grams dry.

  • Use predict() to see if we can use PV curve models to get values collected in the field (predict midday Mpa from LFM):

We’re just interested in ABCO, so just look there:

3.2.1 Simple model - above

3.2.2. Complex model - below:

Since we don’t have Po for the field data, so we’ll have to do the best we can.

3.2.3 Simple model - above

3.2.4 Complex model - above: